Being safe on the internet (was Re: Here we go again - ISP DPI, but is it interception?)

[Ian Batten]

On 12 Aug 2010, at 10:12, Roland Perry wrote:

> In article <4C62F16E.9080405 at bbk.ac.uk>, ken <k.brown at bbk.ac.uk>  
> writes
>
> >A colleague of mine ... [told] the people who took problem calls at  
> >IBM ... exactly what the programmers had done wrong in the source  
> code >of a program for which we didn't have the source code - the  
> error was >produced by an obvious typo in the assembler. Obvious to  
> him that is. >Not me.
>
> I did that few times for DOS4 ! Although (as an OEM who wrote, with  
> my own fair hand, a few customisations) I had the source code for  
> several of the device drivers and utilities, I didn't for the main  
> body of the product. The parts in question were written in Intel  
> assembly code.

Compilers were a lot simpler, though.   Fixing problems without the  
source was easier when compilers generated fairly stereotyped code and  
the object instruction sequence was closely related to the source  
program.   It's a lot harder these days because the relationship  
between source and object is more complex, and the object code has to  
deal with the fact that modern architectures have moved a lot of  
things out of the hardware into the compiler --- a 1960s compiler had  
much less knowledge of instruction sequencing than one today, because  
the issues (ho ho) were different and the hardware designers had done  
a more complete job...

ian